TW201034418A - Cyclic prefix schemes - Google Patents

Abstract

Methods and systems are proposed for transmitting data from a source to a destination via a relay station having multiple antennae. The relay station receives from the source a message containing the data and a first cyclic prefix. It does this using each of its antennae, so producing multiple respective received signals. In certain embodiments, the relay station removes the first cyclic prefix from the received signals, replacing it with a new one. In other embodiments, the relay station removes only a portion of the first cyclic prefix. In either case, the relay station may apply space-time coding to generate second signals, which it transmits to the destination, which extracts the data. Methods are also proposed for estimating parameters of the channel, to enable the destination to decode the data.

Description

201034418 VI. Description of the Invention: [Technology of the Invention] The present invention relates to a system and method for implementing a cyclic first code in a wireless communication channel, the wireless communication channel containing executable information coding, particularly but not completely , analog space time coded relay station. Cross-reference related application

This application is one of two patent applications claiming priority from U.S. Provisional Patent Application Serial No. 61/089,617. This patent application relates to the use of a cyclical first code, while other patent applications relate to techniques for spatial time coding. The techniques described in the respective patent applications of the two patent applications are independent, as any one may be used without the other, however it is equally possible to construct a combination of the respective patents from the two patent applications. The system of application technology. BACKGROUND OF THE INVENTION Wireless relays have been shown to be used to extend communication range and provide a good quality of experience. In a typical communication system that uses relays, the relay station will be time shared between different sources and destinations. In this case, the relay station needs to estimate the angular carrier frequency offset (ACFO) present between each pair of sources and the relay antenna (also indicated by ACFO). The ACFO (represented by ACFO first) present between each pair of relay and destination antennas will be compensated at the destination. Joint compensation from source to relay and relay to destination ACFO is not available in the previous 3 201034418 technology. Thus, the long-term compensation from the source to the relay ACF can only be guessed at the relay station. The prior art _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ It also has the limitation that the transmission channel must be flat fading. This is because inter-code interference (ISI) is generated when the transmission channel has frequency selective fading. The insertion cycle first code (CP) can be used to mitigate ISI. The cyclic first code should preferably have a length greater than the length of the channel impulse response (CIR). A typical implementation of ASTC also requires that the samples in the signal be reordered as the encoding is performed. It is an object of the present invention to provide an implementation The method of looping the first code and compensating for the channel response of the wireless communication channel, which solves at least one of the problems in the prior art and/or provides a useful choice for the public. C SUMMARY OF THE INVENTION J SUMMARY OF THE INVENTION A wireless communication channel is transmitted from a source to a destination by transmitting data from one of the plurality of antennas. The relay station receives the data from the source and the first The message of the first code. The relay station uses each of its antennas to complete the reception, thus generating a plurality of separately received signals. In the first aspect of the invention, the relay station removes the first cycle first code from the received signal, using the new cycle first code Instead of the loop first code, in the second aspect of the invention, the relay station only removes a portion of the first loop first code. In either case, the relay station can apply spatial time coding to generate the first k 201034418 relay station. The first signals are transmitted to the destination of the subject. The s1 of the month of the month is related to the estimated channel parameters to make the destination information available. The present invention can be selectively represented as being matched. An integrated circuit of the above-described level of the present invention. The cyclical first code architecture of the present invention does not require that the transmission channel be flat fading. The drawings are briefly described by way of example only, and the embodiments will be described with reference to the accompanying drawings. 1 is a schematic diagram of a communication channel having one source station s, a relay station R, and a destination station D according to an embodiment of the present invention; FIG. 2 is based on an exemplary embodiment A flowchart of an architecture for inserting a loop first code of a first code architecture 1; FIG. 3 is a flowchart of an architecture for inserting a loop first code of a first code architecture 2 according to one of the exemplary embodiments; 4 is a flow chart of a method for performing channel estimation according to an exemplary embodiment. t Real package mode] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows a communication channel according to an exemplary embodiment. The channel consists of a source station/node 110(S), a relay station 120(R), and a destination station/node 130(D). The source and destination each have only one antenna and the relay station 120 has two antennas 122, 124. The antennas at the source station/node 11〇, the relay station 5 201034418 120, and the destination station/node ι3〇 can be combined to both transmit and receive. Source 110 sends a signal to relay station 12, and relay station 120 sends a signal to destination 130. The signals transmitted from source 110 to relay station 12 and from relay station 120 to destination 130 may include multiple orthogonal carrier frequencies, such as where OFDM modulation is used. This embodiment includes only one relay station 12A. An ASTC split-pass architecture is applied using Orthogonal Frequency Division Multiplexing (〇FDM) modulation at source 11〇. During two consecutive OFDM symbol periods, ASTC codes are applied to the respective carriers. Relay station 120 receives the OFDM signal using two antennas 122,124. It performs simple analog processing of the received signal (such as sampling and storing discrete time signals) and performs encoding. This encoding may for example be a 2X2 Alamouti encoding. The encoded signal is then transmitted to the destination node/station 13〇. In order to avoid interference, the transmission from the source node/station 110 to the destination node/station 13〇 occurs in two phases. Thus the transmission period consists of two phases. In the first phase, the link from the source node/station 11 to the relay station 12〇 is initiated, while the destination node/station 130 remains silent. In the second phase, the link of the relay station 12 to the destination node/station 130 is initiated and the source node/station 11 remains silent. The simulation of this embodiment is implemented using 16 carriers in each OFDM signal. However, this does not limit the number of carriers to 16, and readers with ordinary knowledge will understand that the number of carriers can be different. Alternate embodiments may also apply the ASTC split transfer architecture to a Single Carrier Cyclic First Code (SC-CP) system, where ASTC encoding is also applied to individual symbols on two consecutive SC-CP blocks. Alternative embodiments may also have more than two antennas at the relay station ,0, in 201034418. In this case, antenna selection may be employed. Only two antennas are selected to implement the proposed architecture based on predefined selection criteria (e.g., optimal product channel SNR, etc.). Alternative embodiments may also have multiple relays that relay selections may be performed, with one of the relays being selected to implement the proposed architecture based on a predefined selection criterion (e.g., optimal product channel SNR, etc.). Alternative embodiments may also have multiple relay stations, where coordination delays ASTC^ are implemented at each relay station and coordination delays are applied at different relay stations. In this case, the delay period applied by the different relay stations is a design parameter obtained from the central control. This embodiment may have the advantage that the carrier frequency does not have to be orthogonal so that signal collisions will not occur. Alternative embodiments may also have only one antenna at relay station 120. In this case, one solution is to perform cooperative work using at least two relay substations, each of which is a relay station that can perform information communication with other relay substations. The information channel is implemented between at least two of the relay substations and thus the two fewer relay substations can participate in the ASTC transmission. The angular carrier frequency offset (ACFO) makes the number of carriers W. The channels from source 110 to relay 120 and relay 120 to destination 130 are assumed to be frequency selective multipath fading channels. Let the vectors hs, 1 & hs, 2 denote the channel impulse response (CIR) from the source 110 to the first to second antennas 122, IN of the relay station 12, respectively. Both 115, 1 and 1182 have the size of (L; xl). Similarly, the vectors hu and 匕 are represented as C111 from the first and second antennas 122, 124 of the relay station 120 to the destination 13A, respectively. Both Vd and h2, D have a size of (L2xl). Let f〇, fr, and 7 respectively 7 201034418 be the carrier frequency of the local oscillator at source 110, relay station 120, and destination 130. Then, Mr=f〇-fr and the CFO of the relay 12〇 and the destination 130 are indicated. Source 110, relay station 12, and destination 13A may have separate local oscillators such that their oscillation frequencies need not be the same. This causes the relay station 120 and the destination 13 to receive independent signals in the signal, i.e., Af#Afd. The angular CFO (ACFO) is defined as: Φ, =~{^Γ)τ which also represents the ACFO present on the channel between the source 110 and any of the antennas 122, 124 of the relay station 12, and a represents the presence at the relay station. ACF〇 on the channel between any of the antennas 122, 124 and the destination 130. Γ indicates the period of the OFDM symbol. The scalars ~=(Δ/)Γ and ~=(ΔΛ)Γ are called normalized CFOs, and their magnitudes are limited by μ>〇.5 and |£#〇5. The ACFO may also and/or may be estimated during channel estimation and used to perform compensation at the relay station 120 or at the destination 130. The joint ACFO compensation can be performed at the destination 130, in which case the ACFO compensation needs to be not completed at the relay station 120. This means that the compensation (also + also) is done at destination 130. If the CFOs present in the received signals at the relay station 120 and the destination 130 are not properly compensated, the gain due to the split transfer and the use of OFDM modulation cannot be achieved. It is advantageous when the joint ACFO compensation is completed at the destination 130 because 201034418 maintains the relay carrier frequency for which it is not affected by any source and destination pair. It is also advantageous when the joint ACFO compensation is completed at the destination 13 because it simplifies the implementation of the relay station 120 and reduces computational complexity at the relay station 120. Loop First Code Architecture with Carrier Frequency Offset (CFO) Compensation This embodiment uses two loop first code architectures, the two loop first code frames

The following is described in terms of its associated method of performing carrier frequency offset (CFO) compensation. The cyclic first code architectures 1 and 2 of the present invention allow spatial time coding to be performed at the relay station with minimal signal processing complexity. It should be understood that the word "CP Architecture 1" is also used in this document to refer to the loop first code architecture 1 ' and "cp architecture 2" is also used to refer to the loop first code architecture 2. #环首码 ARCHITECTURE1 Fig. 2 shows a structure for inserting a loop first code according to an exemplary embodiment. Let Xj be a 〇Vxl) vector containing the W symbols to be transmitted during the jth OFDM symbol. This exemplary embodiment of the invention has been described using 2 OFDM symbols 2η^Χ2„+ι), but it is apparent that it is possible to use different numbers of OFDM symbols in the scope of the invention, which It will be apparent to a reader having a conventional fiber. At the end of the experiment 202, the source derives from Xj that the first household with the length "one is not CPj,!". h is the length of the channel impulse response (CIR), for the channel between the first antenna of the relay station and the channel between the source and the relay station's 9th 201034418 antenna, the channel impulse response (CIR) is respectively hs] and Hs, 2 indicates 0. At step 204' the source inserts CPj! in front of Xj and then transmits the resulting sequence of symbols. Let the symbol sequence Xj consisting of W symbols be represented by Χη, where π=〇···^ν一 1 , gp

Xj~[X〇 · · · X". · · 1 ]

Length/^ of the loop first code [Qiao ^ by copying the last A symbol of Xj to generate CPj, l = [% · * · ΧΝ~2] The loop first code CPj, i is inserted in front of Xj, resulting in the resulting The symbol sequence is used for transmission.

The transmission from the source to the relay occurs during the first phase of the transmission cycle. At step 206, the relay station receives L〇>jiXj" at the two antennas of the station. At step 208, after performing time synchronization, the length of the corpse/CPy portion in each 〇FDM symbol is removed leaving ~. If ACF〇 is present, the received signal vector in the time span of 2 OFDM symbols (ie for) = {2 & 2n+l}) is given after removing CPji as: ΓΛ, 1, 2η ri?.l, 2n+l rfi, 2, 2n 1>, 2'2η+ι

Ei=(four)ej((2n+1)((4))+/WrZi)Hsi e;i (+Pl)+Pl)^Z^(^)Hs,2W«x2n

Vfia, 2n Vfi<l2n+1 V/?'2> Vff, 2, 2n+1 fR, i, j represents the received signal vector obtained after CPjj is removed, where ^• and J• respectively represent the antenna index and OFDM symbol period. rRij has size (;vxl). „ indicates the data transmission cycle index. (7VxA〇 matrix is the inverse discrete Fourier transform 10 201034418 for the (IDFT) matrix. (#χΑ〇Circular matrix Hs,! and Hs, the first row of 2 vectors respectively 疋Ι/^,Ι,向量ίχίΛ^Γ and . The vector vR,i,j contains # samples of the additive white Gaussian noise (AWGN) that distort the signal received by the second relay antenna during the yth OFDM symbol. 2 OFDM symbols are transmitted in each cycle, where j is an integer value, then the matrix ~, 丨 丨 "I will have the size (2ΛΓχ2ί/), with LV/?, 2, 2« νΛ, 2 , 2η+1 ^

OFDM symbol index 2, 2 such as +1, 2, +2, ..., 2, +2^1. Suppose the covariance vR, M, where /=1, 2 and j'=2n, 2η+ι, where 1 is selected to be ί±, in the presence of acfo, the compensation ACF 〇 can be performed at the relay station as step 230 . Assuming that the ACFQ is removed or if the ACFG is not present, then in the time span during the (iv) OFDM symbol (ie, for Qiuchuan, the received money vector obtained after removing the CPja at the relay station is given as follows:

Γβ,1,2η Γβ,ι,2η+1 • rJ?,2,2n Γβ,2,2η+1 ·

LhSi S- 'n+l +1 V/?,l,2n vi?4,2n+l VRX2n νβ;2ι2η+1 matrix Yr is counted

Then step 2〇9 is executed at the relay station. In step 209, in order to implement the spatial time compiling at the carrier level, the simple processing is performed and arranged: TY« r (/, /?) y12, /e ^2\, r y22/? rfi, l; 2n ΚΛ2ΚΛ LU Ίη) ==:=: 11 201034418 where As, is a (#xA〇 size diagonal matrix' whose diagonal elements are given by #IR DFT of CIR hs>i, where ί·=1, 2, ie Λ,, + =νϊ^ν^Κ[/^, ο1)<(;ν_Μ:Γ), which can be proved as: Υ = Γ/Ϊ.1.2Π rR.1.2n+l 1_丄[ WMs.lX2n W^AS.!X2„+1 1 y R_[((42,n+i) -((rk2n)J_V![w^ZA!(|)A*,2x*„+1 - W^.Z,v(|)AJ,2X*J+ fl'"' where (2AAx2A〇 size matrix\^,„ is given as: ^ R,n —

Vt,,1,2η νβ,1,2η+1 C (νβ, 2, 2η+ΐ) - C (νβ, 2.2η) Step 209 further includes steps 210 and 212.

At step 210, there is no need to do anything for those elements of the YR that do not require signal sharing (e.g., for yil, R and yi2, R). At step 212, for those elements of the YR that require signal commons (eg, for y2i, R and y22, R), the OFDM symbol sequence is reordered using a mapping function ((.): C(a) = [a (iV-l), a(7V-2), ···, α(0)]τ,

(iVxl) size input vector α = [α(0), α(1 ), ···, a(iV-l)]7'. By storing the signal samples in a shift register and then reading the registers in reverse order, this function can be easily implemented with hardware. This represents a significant advantage because more complex reverse reordering operations (such as inserting, deleting, and exchanging signal samples) do not necessarily need to be completed. Can be proved as: , ([H, „W»x; j*) = =C Tao (W;,), =Ζγ (wearing) W'TW.vW?Tx; = W? Z., (Special - )Λ;,ΤΤχ; =Wy Zv for j = 2/).2/) + I and i ^ 1.2. 12 201034418 where matrix ZN(4) represents the following given one (iVxAO diagonal matrix: Z., ([1 ... And Τ is one (iVx^O permutation matrix, such that T = [^J(1)S7T = ~ Ιλγ is an identity matrix of size #. The operator [·Γ represents a complex number of common rounds. The token [α](1) Represents the vector α cyclic shift / element, where [al_ = [(ί(Λ; —2).η(ι\τ — 1).«(〇)·«(1)· · · , «(·^ ' - • 5) j,, for a particular matrix X, the matrix [χ] (1) is obtained by cyclically shifting each row of vectors Ζ elements. In step 213, length /> 2 (where /&gt ; 2 = B 2_1) The second loop first code (represented as CPW is obtained for each yij R. The heart is the length of the ^, the channel between the first antenna and the destination of the relay station and the second of the relay station The channel between the antenna and the destination, CIR is represented by the vector and h2, D respectively. The first code cPw of each cycle is obtained from the corresponding yijR. Then the first code CPij, 2 is inserted in front of yij R. To generate the symbol sequence LCpij'2yy, R". It is worth noting that yijR is a vector of one (WM) size. At step 214, other necessary steps for ASTC implementation at the relay station are performed. , '' during the OFDM symbol period is transmitted by the ith relay antenna, where W, 2 and the slice ~, ~chuan. The transmission from the relay station to the destination is in the transmission period where the source remains silent. The second phase occurs. In step 216, after the frame synchronization, the OFDM symbols of each of the received 13 201034418 destinations can be represented by the (7νχ1) signal vector rDJ during each yth 〇FDM symbol. ACFO is present and ACFO compensation is completed at the relay station. For the _/={2n, 2n+l} OFDM symbols, the received signal vector at the destination 〇/Vxl) in the presence of ACFO is given as: CP Architecture 1 · = :..ψι.........................f ¢(3⁄4.) w^Txj,,^] + Pft2n

Rfl.2,e = —v7|--~(13⁄4) W^Txy Ten of which the noise vector is given as: gj(2n(iV+P2)+P2)^d Ρΰ, 2η _ ^/| ^ΝίΦά ) [H^pV^j^n + H2,DC (νΛ,2,2η+ΐ)] + vD,2n 6ί((.2η+1)(Ν+Ρ2)+Ρ2)ΦΛ

PjD, 2n+l = 3 Zjv(^d) [H1; DVR 12η+ι -H2, 〇ζ (Vfl.2, 2n)] +vD.2n+l- Selectively 'ACFO also compensates at the destination It is executed as step 240. Assuming that there is ACFO and the ACFO compensation is not completed at the relay station, for the J={2n, 2n+1} OFDM symbols, TD, 2n = βι{2η{Ν+ρ^ρ^ΖΝ(φά) [ej( 2"(Ar+pi)+^)^Hi,i,ZN((/»r)WjA5,iX2n buckle-Zheng 1 oblique Baoqin)(9)M'A+1 +P/?,2n Γβ,2η+1 - βΜ2η+^Ν+ρ^ρ^ΖΝ{φά) [e^((2-+i)(^+^)+Fl)^H1,DZiv(0I-)W^As,iX2n+i _e'qing1 ) + W 'rH2 DZ times) wS ~ (!) Μ Α + P_D, 2n + l · 14 201034418 Η π and H2, D is (iVxA 〇 circulant matrix, where fe, 〇 wy, 1 coffee ~ t) j and [ ΜιΛμΓ is used as its first row vector. (Wxi) vector PD, 2n represents force-induced noise. Assume that ACFO compensation is performed, or if there is no ACFO, 第β(2n) = H1)jDW^A5,lx2n + H2,dW^ Ζαγ ^ Λ·5)2Χ2η+1 -f Ρβ (Nxl) vector PD, 2„ denotes the additive noise ❹ PD given as follows, 2n = Ηιι〇νΛ, ι, 2η + Η2, £. ζ (v^2> 2n+1) + VDi2n, where At the destination (TVxl) vector vD, 2n contains TV samples of AWGN, where the covariance matrix 4^, 2„<2"=<^ heart. Similarly, for the period </_=(2«+1) OFDM symbols, &£>, 2ίΐ.+1

Hl^W^A^^n+l - H2,dW^ Ζ,ν

>2rt.+ l Ρ〇,2η+1 = Η^ί,νΑ,χ^+ι - H2 D ζ (v*R 2 2n) + vD 2n+1.

The ASTC decoding is performed in the frequency domain using the channel state asset (CSI) parameters obtained during the training phase at step 218'. For the symbol period 2π+1}, the defect DFT is performed on the signal vector rD,j. Suppose there is no ACFO' then ~, ^~point DFT is W, = Λι, AlX2„ + λ2, 芸) + a Similarly, the signal received by the destination during the (2n+1) OFDM symbol The iV point DFT of the vector (ie rD, 2„+!) is given as Λέ, 2χ5» ^ W.ivPx>, 2n^l W.jvro.n = — Λ2ί_〇艺九(Ηϊ) 15 201034418 一( 27Vxl) Vector >^, „ is constructed as y'o.n - [(Wa'1*/)*»)7 .(W.Vl*X)^n+l)tf]

Hp + ^A'P/>t2n- ξ s, * X2n^l X2n and X2«+l estimates (expressed as Mao „ and Mao „+1) can be obtained using HP's common roll transposition (ie 〇) »r Υη, '*2„ X 2w+l Η (W vr/} t

Hp* (2iVx2A〇 size product channel matrix and is given as Η ρ

Ai, dA5, i Ζ λ, (labor) A2, dAS2 - Zjv ("ridge" AkAsj Λί, £ > Λ 5, ι product channel matrix 圮 is a known parameter obtained during the estimation of the product channel matrix, and forms a channel state Part of the information (CSI). ^allow a simple linear decoding at the destination because

HpHp -- I_> (|A!, £>|2|As'i|2 + |A2,d|2|A,s.2|2) where ® means performing a Kronecker product . At step 230, the relay station selectively performs ACF0 compensation. For antenna b 2, the received signal vector rR,i,j in the time span _/={2«, 2n+l} during 2 OFDM symbols is: Γβ:1,2η ΓβΛ,271+1 Ri?,2,2n ΓΗ;2.2η+1 e gamma (Λ.则+Ρι)~ΖΛ, (6)e training 2"+1)(狮册(四),Z:v(WHsiWgX2n+1 ej(2"('v +Pl)+Pl)^Zv(0r)Hs,2W^x2n e^(2n+1)(A,+^)+Pl>^ZA.(0r)Hs.2W^x2„+1 VR,l2n Vfi , l: 2n+1 vi?, 2, 2n Vi?i2> +1. 16 201034418 The ACFO bell is a known parameter obtained at the relay station during the ACFO table estimation step 410 and forms part of the channel state information (CSI). By multiplying rR, U„ and rR,2,2„ by 〆2" ((4) conjugate, ACFO also has effects on rR, l, 2n and 2, 2 „ can be eliminated. By rR, l, 2 «+l and rR, 22 +1 multiplied by a 2n+1) (d) (4) Conjugation 'ACFO chopping effect can be eliminated with 丄(6) and above (10). In step 240, the destination is selectively executed ACFO is the compensation. The received signal vector at the destination (ΑΓχ 1) in the presence of ACFO is given as: CP architecture 1: ΓΡ, 2ϊ' ψζ Ζ, νίΦ.) + η2ίϋ ζ (ey wiiT4(+,] + ΡίΛ. \2n x/2 '^χ{Φά\ H2.£> C (13⁄4) W^TxJn] Φρ/>^+1. The noise vector is given by ?0,2„,?0,2„+1. 八匚?〇 is the estimation step in gossip A known parameter obtained at the destination during 420 and forming part of Channel State Information (CSI).

By multiplying rD,2„ and rD,2(4) by the conjugate of 〆 and 'η+1)(Λ^)^ζΛ, respectively, the influence of aCF can be eliminated. Cyclic first code architecture 1 can have the following advantages: That is, only analog domain processing needs to be done at the relay and only linear processing needs to be done for maximum likelihood decoding at the destination. Cyclic First Code Architecture 2 Figure 3 shows the insertion of the loop first code in accordance with an exemplary embodiment. An alternative architecture. Let Xj be a (Μ<1) vector containing iV data symbols to be transmitted during the first 〇FDM symbol. 17 201034418 although this exemplary embodiment of the present invention has It is described using 2 〇fdM symbols (ie ^ and \2 +1), but it is obvious that it is possible to use different numbers of OFDM symbols within the scope of the invention, which is for readers with ordinary knowledge. The words will be obvious. At step 302, the source derives the first cycle first code of length p from the & (denoted as cu, where />= corpse 2 = L; +L2_2.1; is the source no and is represented by hs丨 and respectively The channel between the first antenna 122 of the relay station 120, the channel pulse response (CIR) of the channel of the second antenna 124 of the relay station is the relay station represented by h] force and 'respectively The first step is the channel between the antenna 122 and the destination 130, and the channel pulse response length of the channel between the relay station 12 and the second antenna 124 and the destination 13G.

At step 304, the source inserts CPjl into the sequence of symbols preceding Xj. Then the transmission wheel produces a symbol sequence Xj consisting of one symbol by X", where xj=[x〇...X"...XjV-j] the length of the loop first code cpjl through the copy ~ the last ^ GPj,l = [XjV-/>/...X;V-2 Χλμ] The first code CPja is inserted in front of Xj to generate a sequence of symbols generated by cp words.

From source 110 to happening at step 306,

The transmission of the relay station 120 is received at the two antennas of the station at the first station of the transmission period [CP 18 201034418 In step 308, after performing time synchronization, the CP of each OFDM symbol], 1 part The first six samples are removed leaving a sequence of 2Xj". The first & samples of each -QFDM symbol are distorted by the ISI caused by the frequency selective fading from the source to the relay channel, and the resulting OFDM symbol will have (10) p2) samples. If ACF〇 is present, the signal vector received during the time span of 2 OFDM symbols (ie for {2~仏+ } }) is given after removing CPji: ^R,l,2n Γβ;1.2η+1 .2η Γ/ϊ;2;2τι+1

, _ 2 post (can ej ((4)) (Material > gt; ^ + Qin) fis, 2WgX2n + i yfi.l, 2u Vffi; 2n + 1 1 V «. 2, 2n ^ 2n + l ~; The ((iy+P2)xl) received signal is obtained after the ^^ samples, where i· and ) respectively represent the antenna index and the 〇FDM symbol period. w denotes the bedding transmission cycle index. The (7VXj/v) matrix Wnh is an inverse discrete Fourier transform (IDFT) matrix. The matrix people (where i=1, 2) are given as follows ((10)(5) which matrix parameter Ssi = H5!i(iV - P2 + 1 : iv,:)' , L Hs; ij, where the citizen is one ( #><Λ〇Circular matrix, where Mf as its first row inward' and Hs/AT-ZVID represents the last P2 column vector of Hsi. The vector of size ((N+Z^xl) contains the influence signal匕, AWGN 共 The covariance matrix of the noise vector of the relay station is given as 成 〇 + + / / 2) Selectively, in the presence of ACFO, ACFO long compensation can be performed as a step in 19 201034418 relay station 330. Assuming that the ACFO is removed or if there is no ACFO, then in the time span of the 2 OFDM symbols, ie for)={2«, 2«+1}, the CPjj of each OFDM symbol is removed. The received signal vector obtained after the first A sample of the part is given as follows: ·f where l,2n fi?,l,2n+l Hs,iW^x2n HS)1W^x2n+i -L H5,2W ^X2„ H5,2W^X2„+i 丁_ ^R.2,2n V^(2,2n+1

Then step 309 is performed at the relay station. At step 309, in order to perform spatial time coding at the carrier level, processing is performed. The matrix YR is calculated and arranged: yii, /? yn, R 1 ΓΛ, 1, 2η _yn, R y 22, R _ = Έ - U_

The operator [I represents a complex conjugate. The spatial time coding implemented may be, for example, space time block coding. Step 309 further includes steps 310 and 312.

At step 310, there is no need to do anything for those elements of the YR that do not require signal conjugation (e.g., for yu, R and yi2, R). At step 312, for those elements of the YR that require signal commons (eg, for y2i, R and y22, R), the OFDM symbol sequences from the respective frames are reordered using a mapping function ((.): C(a) = [a(iV - l), a(iV - 2), · · · , α(0)]τ, (TVxl) input vector ΑΤ = [ή((9),β(1),.··, ί/ (τν-1)]τ. By storing the signal samples in a shift register and then reading the registers in reverse order, this function can be easily implemented with hardware. 20 201034418 Thus the mapping function ((_) can be re Expressed as: ζ{α{η))=α(Ν -nl) where α(«) represents the „th symbol of the OFDM symbol sequence α. Then the _time domain signal is conjugated in each reordering symbolΓ (α(8)) is executed, where «=0... A^-l. Let 6 = ((a). It can be proved that after the signal conjugate is completed, B={fi(9)...5( /:)...β(ΛΜ)} is substituted into the corresponding frequency domain sequence of b

That is, the phase shift conjugate sequence. The following discrete Fourier transform (DFT) characteristics exist: • Linear: 〇^(«)+Bu(")<=>"^:(moxibustion)+£^(moxibustion); • Cyclic shift Jc (〇 i+m)/v) <=> wfX (4); • Symmetry / ((-n)w) <=> / (magic

x(n) and :Kn) are used to represent the time domain sequence, phantom and y (the corresponding frequency domain sequence, WDFT size, and w"=exp \ in step 314, for other necessary steps in the ASTC implementation of the relay station) 〇^^YR4yJ=[y;^ is transmitted during the yth OFDM symbol by the first relay antenna, where Bu 1, 2 and _/= 丨 2 «, 2n + l}. From the relay station The transmission to the destination occurs during the second phase of the transmission period in which the source remains silent. At step 316, after the frame synchronization, the destination removes the long-corrected view first code from each received OFDM symbol S. Each 21 201034418 OFDM symbol after the code removal will be iV sample lengths, and may be represented by the signal vector rp,j for each and every 〇FDM symbol period. Assuming that ACFO is present and ACF〇 compensation is completed in the relay, for the period of >/={2«, 2«+1} 0 corpses 01^ symbol, in the presence of gossip? The ground (ATxl) received signal vector is given as: CP architecture 2: r/, '& : 3 , off s.iWgx& + iia.GeWgTxL Hj + ^./ ί +1Η i+/3⁄4

= ^ ^lJ:>^K^2n+l - H2sDG5.2W^Tx*J where the noise vector is given as ~ ei(2n(N+P2)+P2)^d ^D'2n ~ ^ Zjv (0d)RCi) [Uli£)Vfla!2n + U2,DC (vR,2,2n+l)] + VD,2n . ej((2n+l)(AT+P2)+P2)0d

Pd, 2ti+1 = 'Zjv(</>d)RCp [Ui^Vfl.l^n+l -H2,DC (v^2,2n)] +vD,2n+l· Selectively, The ACFO nine compensation can be performed at the destination as step 340. Assuming that ACFO is present and ACFO compensation is completed during the relay, for the 2n+1+1 OFDM symbols, Γρ, 2η = βΗ2η(Ν+Ρ2)+Ρ2) ίφ, +φ,)ΖΜ + W^X2n + 7nH2,0GS,2 (W^X2n+1)* + P〇,2n Γ0,2η+1 = εΛ(2η+1)(Ν+Ρ2)+Ρ2)(Φ<ι+φ,.)ΖΝ^ Φα + W^x2n+1 -7„e^1^H2^G5,2 (W^x2n)*] +pD,2rt+1, where Ha and H2, D are (iVxA〇cyclic matrix, where 〇1\ (>1丄2) "And 2士夂"[/^,〇·!^!" as their respective first row vectors, and scalar 乂 and 匕201034418 are defined as follows: 〇ίη =: ^( 2n-\-l)p1<j>r 7n = ej((4n+2)(iY+P)_2nP1-l)^r Optionally, joint ACFO compensation can be performed at the destination as step 350 to compensate ACFO and IX. Suppose ACFO compensation is performed, or if there is no ACF 〇, for the _/·=(2η) OFDM symbol period,

^, 2n = Hi, 〇H5, iW^x2n + H2tDG5, 2(W^x2n+a)* + pD2n, (ΛΤχΛ〇 matrix gs 2 is given as

Gs,2 = c{&s2(l: N,:)Y The mapping of the two-dimensional matrix A ((·) is calculated as follows: ···., where % .··, ΛΓ 2, then A has one Size (Α^χΛ^) and A = [% , ar2 is a (Wxl) vector, where /=ι, 2, ((4) round 1) '%)·,·., 'J. (ΛΓχ1) Vector ^2„ is given as

P^n = Rcp [UltDVflili2n + U2; DC fe, 2n+1)] + VD, 2n, where the position v〇2" contains the AWGW received at the destination, for 2, the moment drop is one (〇v +P2)x(iV+P2)) 蚝波力兹矩阵式, where ^, 〇_ ^ is its first row vector and [&(11)〇|χ(Ν+ρ dJt is its first column vector. The matrix Rcp is a (#x(w+p2)) matrix defined as follows, I^cp - [ 〇Nxp2f ΙλΓ similarly received at the destination (iVx 1) during the (2n+1)th OFDM symbol The size signal vector is obtained as rD, 2n+1^ η1<Βη3,^χ2η+1-Η2^5,2^χ2ηγ + ρΠ2η^ PD,2n+l - [υΐ!Ζ?νβ)1,2η+1 ~ U2ii? ζ (v^2, 2n)] + V£, i2n+i. 23 201034418 In step 318, the destination performs ASTC decoding in the frequency domain using channel state information (CSI) parameters obtained during training. During the symbol period j={2n, 2η+1}, the defect DFT is only performed on the signal vector w, and the generation of a (2Nxl) vector is constructed as

X2«4-t

^A'Pt\2n (Wa-correction)* The signals rD 2„ and rD 2/>+ can be expressed as rD, 2n = Hi; £)H5>1W^X2n + H2,r»G5,2 ( W^X2„+l) + Γ〇, 2η+1 = H1)£)H^iW^X2ri,+ l ~ Η2:£)〇5:2 (W^X2r7.) + Pz), 2n+li matrix <^, 2 is defined as one of the following forms (7VxA〇 matrix: 〇5.2 = c(H;, 2(l:iV,:)). It can be seen that the matrix Gs, 2 can be arranged by one ( iVxA〇circular matrix cs>2=w^A4s, 2wN row vector is obtained. Using this knowledge, it can be proved that

Substituting (7VX1) size vector Η2^,2«χ2„+1)4 into Gs,2, h2,dgs,2 (W^x2n+1)* = W^A2,D(WNG5.2W^) W^ (W^x2n+1)* =w>2,D ((zw (芸))i2 a*s'2t) WN (Wgx2„+1)* =w^(ziV (芸))2a*s,2ttx ; +1 = W 炅 (ZiV (P)) A2MA * s ^ 2n +1 since TT = 1N. 24 201034418 Similarly, the (Nxl) vector h2dgS2 (w» can be expressed as H.. „Gs, _, (W^X,,,)* = W» (zy Α,.Γ,Λ^ Substitute the representation of H2, DGS, 2«^„+1)1 H2.dGs,2«x2„)* into ~„, (WjvrD, 2n)T, T Ai, dAs,i (Ziv(^)) 2 Α2βΑ*32 X2n 1 WjvPz>, 2n .-(Zjv (-f)) 2 K,d^s,2 .X2n+1 X. (W; vpA2fl+1). The estimates of ~ and A +1 (expressed as ϋ +1) can be obtained using Hp's common transposition (ie H?)

Χ2Π X 2n+l

Ky〇,n Η ^ΝΓ0,2η (w^r^ 2n+11

Hp is a product channel matrix of 2A^x2A〇 size and is given as

HP Λι,〇Λ5,ι (Zjv (^))1,2 ^2.dA*S2 -(ΖΛΓ (~lf ))L2 Λ2,0Λ5,2 Λΐ,ϋΛ5·,1 Product channel matrix Hp is in the product channel A known parameter obtained during the matrix estimation and forms part of the channel state information (CSI). Hp allows for a simple linear decoding at the destination because HpHp = Ij · + where ® indicates the execution of a gram of kinky product. It can be seen that this embodiment has an advantage because there is no need to complete any sample reordering at the destination. In the presence of ACF0 and the joint ACF0 compensation is completed in step 316, the ASTC decoding in the frequency domain can be performed using step 320 instead of step 318. 25 201034418 At step 320, the (2iVxl) vector will be y〇fr (e-i<M«+ft)+nM«,)w_v35;e{^/)rD-!S,y,(e->! (3⁄4.+lH^^4-ft')+·ί3⁄4ι(¢/)wΛ·z3⁄4(^/)r/5^+.,), X2n^-i

In (|r))

Hr ten ih X2« X2n-i-l + a„

The scalar quantity is defined as an — εΛ2η+ι) ριΦτ ry _ e-j((4n+2)(7V+F)-2?iPi-l)^r which also represents ACFO. The value may also be estimated at the relay using ACFO estimation step 410 and then forwarded to the destination. And an estimate of A +1 (expressed as heart and f 2 „ +1) so that it can be obtained using fiF,

*2re -H

.......y.T —«FJD.t?· x2n+l This means that the fiF is transposed. fiF has the following characteristics: ftgiilF = 12 Θ (|Ai,d|2|As,iJ2 + |A2,〇|2|As,2p). where ® means performing a Klonike product. Thus sF allows for linear decoding of a sample at the destination. At step 330, the relay station selectively performs ACFO compensation as well. In the time span of 2 26 201034418 0FDM symbols, the received signal 2, 2; (where antenna h ° 2) is Γλ,1> TR, L2n+l _ ri? 2,2n ri?,2.2n+le,((2n+,)(-V+m/,I;"Z^(0r)HSl, νΛ.1,2η V» i2n+l 1 ' ^Λ,1 , 2η Vfl,i,2n+1 h,2,2n 々 brother 2.2n+l to +1+1 tiger x2nxsn * 〇nr vr / tablet ACFO is obtained in the ACF moxibustion estimation step 4i〇 -6 knows the parameters and forms the channel state information (CSI) κ through multiplying by 6 ♦ husking (4), ACF〇&pairs; factory ", 2" effects can be eliminated. Multiply by one 2 „+丨)(Λ/+/1)+/^Ζ (ACF0 & the effects of ~, 12»+1 and ^"+1 can be eliminated. In step 340, the destination selectively performs ACFO compensation. At ACFO, the destination (#χ1) received signal vector is given as CP architecture 2: ', ,, f ] eJCSnt.Vi-PaHFa).^ (3⁄4} [ϋ!IW^X2« i H2lnC5s , 2wj?TxJ, ^^ f pXJki ί 4-1 )(N+ ) +73⁄4 1, where the noise vector is given by and the heart n+1. acfo A is one obtained during the ACF 〇^ estimation step 420 Known parameters and form a negative channel state Part of sil(CSI). By multiplying L, 2" and rfi2 +1 by the conjugate of ^ Ζ 2 "+ΐΗΛί+/>2Κί>2ΜΆ(Α), the effect of ACFO can be eliminated. 350, the destination selectively performs joint acf〇 compensation. South to the hUn, 2«+1} OFDM symbols, received at the destination (TVxl 27 201034418 signal vector is given as ΓΑ2" = + ,r ) + 7„h2,dG5,2 (W«x2„+1) + Ρ〇, 2η r0,2n+1 = -7/ton 2,, Gs, 2(Wh2„r]+'t+i, where The noise vector is given by the heart and the heart 2„+, and the scalars \ and ^ are defined as follows: 〇tn — ei(2n+l)Pi^r = g~j((4n+2)(7V+P _2nPi —l)0r It can be seen that there are ACF 〇 and ACFO in the signal vector b2n & rD2n+i. ACFO and ACFO can be added and can be represented by ACF amp & = color + long. ACFO~ is a part of the known parameters obtained during the joint ACF〇 (first + also) estimation step 43 0 and forms channel state information (c s work). The effect of acf〇~ can be eliminated by multiplying ^ and ~, 2"+丨 by the conjugate of W2"(4)(4)W A(么+也) and 2_(_>2)(_2池+也) respectively. The first code architecture 2 may have the advantage that only analog domain processing needs to be done at the relay, and only linear processing needs to be done for maximum likelihood decoding for each carrier at the destination. It is superior to the prior art because the relay station does not have to perform time sharing between different sources and destinations. In the prior art 'when the joint ACFO compensation at the destination is not completed, the relay station will need to target each pair of sources and purposes. Estimate and compensate ACFO. This may complicate the relay station hardware and may increase the computational complexity required by the relay station. 201034418 Channel Estimated Channel Estimation The actual information is transmitted according to the method shown in the flowchart of Figure 2 or Figure 3. Previously performed. The purpose of the channel estimation is to obtain a set of Channel State Information (CSI) parameters for decoding the signals received at the destination, and optionally at the relay station or destination compensation angle. Wave frequency offset (ACFO). Estimation will next be described with reference to FIG. 4, according to the channel to be this exemplary embodiment.

At step 402, a simple pair of pilot symbols (or training symbols) is used to enumerate the product channel. A pilot symbol pair consisting of [χ2„χ2η+1] is transmitted from the source during two symbol intervals. The leading symbol is defined as A» = a χϋίΐ+ι = where α is a (Wxl) vector, such that |α(〇| = 1, where ί=0, 1,..., iV-l and α have a low peak to average power value ratio (PAPR). η represents the data transmission cycle index. The estimate of the ACFO table at the relay station is Step 410, the ACFO table is estimated to be executed at the relay station for the loop first code architecture 1 or 2. Assuming the known value of the pilot symbol [12 „\211+1], the ACFO table at the relay station can be easily estimated as the CP architecture 1 : I = ^(~TR.1.2nrH.U2n+l) + + —— 2(iV + F) ...................... CP Architecture 2 : 29 201034418 - ^(—;1.2n+l) + ^ 22/( ?Λ<2ι2η+1) <pr ^ 9 ^ jV _jT^pjj ~-, operator 4.) returns the phase of (·), Thus the scalar is returned in the interval [-ππ) a Ίτΐε. It is worth noting that the CFO can be estimated and thus the maximum normalized CFO is given the following bounds Ν η, k:r| <〇·^νΤρ<

The above limitations can be overcome by designing appropriate pilot symbols. It is estimated that the ACFO value can be relayed to perform compensation during the actual information transmission. The cyclic first code architecture 2 may have the advantage that ACF〇 compensation in the relay is not necessary because the joint ACFO compensation can be performed at the destination. The ACFO纥 estimated station at the destination is for the first code architecture. At step 42, the ACFO is estimated to be executed in the first or second. " ‘ at the destination

For loop first code architecture 1, in the presence of ACFO (#xl) the received signal vector is given as CP architecture 1: 2 y/2 + H2J) < (H^l λ'Τχ2η+ι] + ΊΪ ;?«J +P/>, 2„+1· where the noise vector is given as Ρ^, 2« eJ(2n(JV+P2)+P2)0d z P〇, 2n+l = ·----- ----^ 72 '

Ar(0rf) [H3iX)V^tlt2n + H2i〇 C (v ^2n+1)]+v〇 2n z

:,DC (V2,2„)] + VA2„+1_ 30 201034418 Similarly, for CP architecture 2, the (Nxl) received signal vector at the destination is given as CP architecture 2: rn, 2it TM 2; vifA /} [HE!〇H5jW^X2,i 4- H2.〇〇iV2W^Tx2n_i.1] + P〇:2n ij -Ji, 4 l!f λ + Ρ, w /J ( r«;—> «.h = ·Ζ;ν(^) H2.dGSi2W»Tx;„] + Po:2ji+i. where the noise vector is given as 6](2η(Ν+Ρ2)+Ρ2)φί1 Ρϋ, 2η - ZAr(^)Rcp [Ui^Vfi)1^2n + υ2,β ( (vjj,2,2n+l)] + VD,2n

Ej((2n+l)(iV+P2)+P2)^ P〇,2n+l — ^Ν(Φά)^αρ [Ui^V^i^n+1 - H2,D ( (^2,2n )] + VD, 2n+l· The following estimates apply to the loop first code architecture 1 or 2. Use the leading symbols transmitted by the source χ2„ =or and χ2„ =-α, assuming = (NxA〇 matrix W/Yes The inverse discrete Fourier transform (IDFT) matrix. Therefore, the vector β1β//51ΐ<χ2„ can be expressed as =Ah/-». where (Νχ(尸+1)) 波波力兹矩阵 A is given as A = [[b]lu,.[b]n).··· .[b](,>,] and ((P+l)xl) vector heart is

~ ^1.0 * t^SM where * indicates a convolution operation. Let us assume. Vector H2DC(lTs'2)d = B heart' where (#x(P+l)) 波波力兹 matrix B is given as B = [£)(-11+1), [£|(-^ 1+2),...,[0(-^1+^+1) 31 201034418 where / =((4) and ((corpse +l)xl) vector heart is hp - D * C (^5,2) * similar Ground, vector H2i)Gs〆 can be expressed as H^G^d = Civp, where (#x(尸+1)) 波波力兹矩阵 C is given as C—[幻(-2L2+1)[ Illusion (—2Z/2+2) · · _ [magic (a 2l2+P+l)

Bring the above results into rD 2„ and 2„+1 and remember rD 2„ and rD 2„+1 are the received signal vectors at the destination's leading symbols, ~, 2„ and & 2„+1 Can be rewritten as CP architecture 1: ^D:2n ^D.2n+1 s/2 ^ j ((2n 4-1) {N 4- P-2) 4·· P-ι} φκ\ — Ζ .γ(4/) [A, B] hF + pD, 2n Zv(drf) [-A, -B] hf + Pd,2,t+i CP architecture 2 : ^D:2r, [A. C] Hf + P〇,: 2n Ϊ*£>, 2η41 Ζχ(side-A·-CjhF + h 2nf 1 '

T

Where (2(P+l)xl)vector/iF hTP, hp makes % a (2#xl) vector constructed as follows

Qn — [rD, 2m rD, 2n+l] ~ F(0c;) h/? + en, where ~=[/ΊΑη+1Γ is used for CP architecture 1, for CP architecture 2. (27Vx2(P+l)) matrix /^d) is given as CP architecture 1: 32 201034418

Z full) 〇, Vx, VA -B 〇Vx^ ej^^zN^d) —A —B CP architecture 2 : F(^) Using Zha, a maximum likelihood (ML) ACFO and CIR estimator can be used planning. ACF0 is estimated to be

Secret)...〇ΛΜ A -c- 〇ΛκΛ' — (叙) _ _ -A -c _ heart=argmjn(qf[I2iV —F(4)(FM)F(4);)-1F ugly (4)]q„)· A grid search is estimated. It is worth noting that for the pilot symbol used in step 402, the matrix (4)F(4) is given as ¥Η(φ)¥(φ) = Ν12{ρ+1). This means (4)F(4) The inverse does not need to be calculated for each search point. Therefore, ACFO can be estimated as Φ ά irgmjn (qf

l2iV

N F(4)F ugly (4)

It can be seen that the ACFO estimate at the destination is not computationally required because the inverse matrix calculation does not need to be done for each search point. Once ACFO 4 is estimated, the vector heart can be estimated as b = 士F(,)m)q„. The joint ACFO (also +) at the destination is estimated at step 430, and the joint ACFO (when + first) is estimated at the destination. The implementation is performed for the loop first code architecture 2. The received signal vector at the destination (Nxl) (without ACFO compensation in the relay) can be written as follows: 33 201034418 rA2„ = #2η(ΛΓ+Ρ2)+Ρ2) (know +~)ΖΛ他+ *)卜:ftwHwWgxh + (\<x2„+1)* +P〇:2n r0,2n+i = 6^2η+1^Ν+ρ^+ρ^+^ ΖΝ(φά + φΓ) [anejPl^H1, DHs, iW^x2ll+1 -7ne^U2, DGSt2 (W^x2n)*] +pD.2n+1, where Η1β and Η2β are respectively (;VxA〇circular matrix , where ~(-<〇[☆,〇丨X(N-L2)r and Α(-(〇[/^,〇丨_J as its first row vector, and scalar and h are defined as follows:

二^j(2n-\-l)Pi4>r _ -j((4n+2)(N-^P)—2nPi — l)4>rq« = Κ,2η, rUr =F^OilF + en, Which (27Vx2 (corpse +1)) matrix? (. life) is given as

Ej(2n(iV+P2)+P2)^ Z』V(泠/) 〇ΝχΝ a„A _7nC V~2 l 0NxN e^p^Z^f) \ -a„e 淋 A -lnejP^C

Yin (</>/,*) and bogey is a (2(P+l)xl) vector constructed as follows: hF =

*hs,i

(&2,D * C (3⁄4 Τ' τ FH (φί,φΓ)Έ'(φί,φΓ) = iVI2(p+i). The necessary maximum likelihood estimate is obtained as Φί

~¥(φ,φΓ)¥Η(φ,φΓ)

34 201034418 and & It is estimated that although the invention has been described in detail, many variations within the scope of the invention are A "疋能能#" which is intended for readers with ordinary knowledge. It will be obvious that the brother, for example, can change the parameters such as the number of carriers, the length of the cyclic first code P, the p1, the coding technique used at the relay station, or the T旳 mode in which the product channel is estimated. Can also make

The design of the leader symbol changes. Going again, the CP architecture does not necessarily use spatial time coding, but other forms of coding can be used. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic diagram of a communication channel having one source station s, a relay station R, and a destination station D according to an embodiment of the present invention; FIG. 2 is a diagram according to one of the exemplary embodiments. Flowchart of the architecture of the first code architecture 1 for inserting the loop first code_; FIG. 3 is a flow diagram of the architecture for inserting the loop first code of the first code architecture 2 according to one of the exemplary embodiments; FIG. A flowchart of a method for performing channel estimation, in accordance with an exemplary embodiment. [Description of main component symbols] no.··Source station/node or source node station/destination/station/source 202~240, 302~350, 402~430·.· 120... Relay station flow steps 122, 124· ··Antenna 13〇·.. destination station/node/destination node/ 35

Claims (1)

201034418 VII. Patent Application Range: 1. A method for transferring data from a source to a destination by a relay station having multiple antennas, the method comprising the steps of: receiving by using the plurality of antennas a message sent by the source, thereby forming a plurality of respective received signals, the message including the data and including a first cycle first code; removing the first cycle first code from the received signals; encoding the received signals, Forming a plurality of second signals; inserting a second loop first code into the plurality of second signals; and transmitting the plurality of second signals having the second loop first code to the destination. 2. A method of transferring data from a source to a destination by a relay station having a plurality of antennas, the method comprising the steps of: receiving a message transmitted by the source using the plurality of antennas, Forming a plurality of respective received signals, the message including the data and including a first cyclic first code; removing a portion of the cyclic first code from the received signals; encoding the received signals to form a plurality of second And transmitting the plurality of second signals having the remainder of the first code of the loop to the destination. 3. A method of transmitting a message by a range as described in claim 1 or 2, wherein the received signal comprises a first received signal from a first one of the antennas of the relay station, and The second received signal of the second antenna of the antennas of the relay station, the step of encoding the received signals, the step of combining the first received signal and the second received signal to form the plurality of The second signal. 4. The method of transmitting a message by means of a sub-range as described in claim 3, wherein the step of encoding the received signals further comprises the step of obtaining a conjugate complex of the first received signal. 5. The method of transmitting a message by means of a sub-range as described in claim 3 or 4, wherein the step of encoding the received signals further comprises reversing a sequence of symbols in the first received signal The steps. 6. The method of transmitting a message by a split range according to claim 5, wherein the step of reversing the sequence of symbols in the first received signal further comprises storing the first received signal in a shift The step of reading the shift register in the register and then in reverse order. 7. The method of transmitting a message by a splitter according to any one of claims 3 to 6, wherein the step of encoding the received signals further comprises the step of negating the first received signal. 8. A method of relaying a message as described in any one of the preceding claims, wherein the received signals are encoded by spatial time coding to form the plurality of second signals. 9. A method of delivering a message by a split range as described in claim 8 wherein the spatial time code is a space time block code. 10. The method of transmitting a message by a splitter according to any of the preceding claims, further comprising the step of performing carrier frequency offset (CFO) compensation on the message at the relay station. 11. The method of any of the preceding claims, wherein the method of transmitting a message 37 201034418 further comprises the step of performing CFO compensation on the message at the destination. 12. The method of transmitting a message by a splitting according to any one of the preceding claims, wherein the relay station having a plurality of antennas is composed of two relay substations, each of the relay substations having at least one antenna, the two The relay substations are grouped to perform information passing between each other. 13. A method for transferring data from a source to a destination by a relay station having a plurality of antennas, the method comprising the steps of: A (i) the source is in a message containing the first code of the first loop Transmitting the data; (ii) the relay station: receiving the message using the plurality of antennas, thereby forming a plurality of respective received signals; removing the first cyclic first code from the received signals; encoding the received signals Forming a plurality of second signals; inserting a second loop first code into the plurality of second signals; transmitting the second signals having the second loop first code to the destination; and (iii) The destination receives the second signals and retrieves the data. 14. A method of transferring data from a source to a destination by a relay station having a plurality of antennas, the method comprising the steps of: (i) transmitting, by the source, a message comprising a first cycle of the first code (ii) the relay station: 38 201034418 receiving the message using the plurality of antennas, thereby forming a plurality of respective received signals; removing a portion of the cyclic first code from the received signals; encoding the received signals Forming a plurality of second signals; transmitting the plurality of second signals having the remainder of the loop first code to the destination; and (iii) the destination receiving the second signals and extracting the data. 15. The method of claim 1, wherein the method further comprises the step of: performing CFO compensation on the message at the destination, thereby compensating for the CFO from the source to the relay station and From the relay station to the CFO that exists at the destination. 16. A method of estimating channel state information comprising a source, a relay station having a plurality of antennas, and a wireless communication channel of a destination, the method comprising the steps of: first antenna of the antennas of the relay station Receiving a first training signal from the source; a second antenna of the antennas of the relay station receives a second training signal from the source; the relay station: (i) spatially encoding the first training signal and the second Training signal encoding to form a plurality of relay training signals; and (ii) transmitting the plurality of relay training signals at at least one third symbol interval; and the destination receiving the relay training signals, and from the Waiting for the relay training 39 201034418 training signal to estimate the channel status information. Π 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 The step of the offset (CFO) parameter. 18. The method for estimating a wireless communication channel channel state information according to claim 17, wherein the step of combining the first training signal and the second training signal performs a total of one rotation on the first training signal. Set. 19. The method of determining channel state information of a wireless communication channel according to any one of claims 16 to 18, wherein the first/training signal includes a first pilot signal, and the second The training signal includes a second pilot signal, which is an apricot of the first pilot signal. 20. As described in any of the 16th to 19th patent applications! The method of squeezing the channel state information of a wireless communication channel, the input of the data must include a step of receiving a combined training signal at a destination station, the combined 丨# signal comprising the plurality of relay training signals and a combined noise signal. 21. A method of estimating a wireless communication channel force channel state information as described in claim 20, wherein the combined noise signal comprises a second carrier frequency offset (CFO) estimated by the destination. 22. A method of estimating channel state information for a wireless communication channel as described in claim 21, wherein the second CFO is estimated by a maximum likelihood estimate. 40 201034418 23. A method of estimating channel state information for a wireless communication channel as described in claim 22, wherein the maximum likelihood estimate is a grid search. 24. The method of estimating channel state information of a wireless communication channel according to any one of claims 17 to 23, wherein the relay station is configured to perform a CFO compensation using the first CFO parameter. . The method of estimating channel state information of a wireless communication channel according to any one of claims 21 to 23, wherein the destination station is configured to perform a CFO using the second CFO parameter make up. 26. The method of estimating channel state information of a wireless communication channel according to any one of clauses 21 to 23 of claim 17, wherein the first station is estimated at the relay station The CFO parameters are forwarded to the destination, the destination being configured to perform a joint CFO compensation using the first CFO parameter and the second CFO parameter. 27. A method of transmitting data from a source to a destination along a wireless communication channel by a relay station having a plurality of antennas, the method comprising the steps of: as in claim 16 through 26 Any of the methods described for estimating channel information of the channel; transmitting a message containing the data along the wireless communication channel; and extracting the data from the message using the estimated channel information. 28. A relay station for transmitting data from a source to a destination and having a plurality of antennas, the relay station being configured to perform the method of any one of claims 1 to 12 . 41 201034418 29. A system for a source, a destination, and a relay station for transmitting data from the source to the destination and having a plurality of antennas, the system being configured to perform as claimed The method of any one of items 13 to 27. 30. An integrated circuit (1C) for a relay station configured to transfer data from a source to a destination, the relay station having a plurality of antennas, the plurality of antennas being configured to receive a message transmitted by the source and thereby forming a plurality of respective received signals, the message including the data and including a first cycle first code, the 1C comprising: an interface configured to receive the plurality of antennas from the plurality of antennas And a plurality of respective received signals; a first processing unit configured to remove the first cyclic first code from the received signals; an encoder configured to encode the received signals by spatial time coding, Forming a plurality of second signals; a second processing unit configured to insert a second loop first code into the plurality of second signals; and the interface is further configured to have the second loop first code The plurality of second signals are sent to the plurality of antennas for transmission to the destination. 31. An integrated circuit (1C) for a relay station configured to transfer data from a source to a destination, the relay station having a plurality of antennas, the plurality of antennas being configured to receive a message transmitted by the source and thereby forming a plurality of respective received signals, the message including the data 201034418 and including a first loop first code, the ic comprising: an interface configured to receive from the plurality of antennas The plurality of respective received signals, a processing unit, configured to remove a portion of the cyclic first code from the received signals; an encoder configured to encode the received signals with spatial time coding, Forming a plurality of second signals; the interface is further configured to transmit the plurality of second signals having the remainder of the cyclic first code to the plurality of antennas for transmission to the destination. 43